Calculating your feedrate to get awesome prints!

Use science to increase the quality of your prints! By finding the actual feedrate of your machine, and dropping that into skeinforge, you can get much better prints!

The DC motors used in the MakerBot are the cheapest possible solution to getting the torque you need to drive a MakerBot but they are DC motors so they don’t all run at the same speed. Speeds can vary from 17mm/sec to 32mm/sec. That’s a huge range and so it makes sense that each bot needs to be tuned. With the current default settings in ReplicatorG 23, we’ve been able to get great results by calculating the actual feedrate of the motors, changing that setting in Skeinforge from 32mm/sec to the actual feedrate and rocking great prints.

One of the easiest ways to get frustrating prints is by having a feedrate that is too fast. If you’re machine extrudes at 17mm/sec and you’ve got it set at 32mm/sec, the filament will get stretched as it comes out of the nozzle and stretched ABS wants to shrink and so you’ll get cracks. If you stretch the filament, you’re objects will get much more shrinkage in prints. My personal MK5 plastruder extrudes at 22mm/sec.

How do you calculate your feedrate? Here’s a way I’ve come up with that works.

1. Raise your extruder nozzle 100mm off the build platform. You can use the control panel to make this happen… or a ruler!
2. Get the plastruder to extrusion temperature. We keep it locked at 220.
3. Start the plastruder by setting the speed to 255 and pressing the forward button. This will make plastic come out.
4. Get a stopwatch ready. Most phones come with stopwatches these days.
5. Use tweezers to pull away the plastic from the nozzle as it’s running. Start the stopwatch at the exact same time.
6. Stop the stopwatch when it hits the platform.
7. Repeat 10 times. Ignore any results that vary widely and average them by adding them up and dividing by 10. You should have a number between 3 and 6 seconds.
7. Divide 100 by your number to get the mm/sec.
8. Drop that into skeinforge under the speed tab and try a print!
9. Win!

Here is another skeinforge tip that we’ve been playing with that seem to be very helpful in dialing in good prints. Up your travel feed rate to something like 40mm/sec. It’ll make the whiskers between objects much thinner. If you’re feeling adventurous, increase it to 100mm (untested!).

Tagged with 21 comments

21 Comments so far

  • Ryanp
    December 22, 2010 at 2:55 pm

    This really bookends with the speadsheet I setup to calculate the other major extrusion parameters.

    It’s getting a bit messy with people editing it. If anyone has a better idea on how to set this up, send me a message at rplumley (ummm)

  • Feilen
    December 22, 2010 at 4:32 pm

    Awesome idea! Calibrated mine to exactly 24.18 mm/s, though I probably have at least a little error, that’s way closer than the 28 I had before.

  • Per Östergren
    December 22, 2010 at 4:48 pm

    Excellent tip! Any advice on how to make the raft come off more easily? My prints require quite an effort with my exacto knife. Saw a video from one of the makerfaires, and on that the newly printed part detached from the raft like a Velcro. Am an utter newbie, have a brand new thingomatic, #17.

  • Zh4x0r
    December 22, 2010 at 5:24 pm

    wow, I must have a really fast plastruder!!! after calibration, I have 41.5 as my feedrate! I’m just concerned that that may be too fast for the steppers with the ABP, as it’s heavy and doesn’t like speeding up and slowing down that fast… Thanks a ton for this post tho!! hopefully i can get the machine 100% dialed in!!


  • MakerBlock
    December 22, 2010 at 5:41 pm

    @Per Ostergren: You can play with the raft settings quite a bit. Your best bet to creating a raft that comes off easily is a thick bottom later, so that the raft will stay flat, and a very thin sparse second layer – so that the pieces will pop off. Congratulations on your new Thing-O-Matic!

    @Zh4x0r: Some Skeinforge calibration guides suggest setting the feedrate and travel rates to the same value. Even with the heavy Automated Build Platform, you shouldn’t have a problem cranking the travel speed up to 41.5mm/s. Now, 100mm/s might get pretty sketchy…

  • Dave Durant
    December 22, 2010 at 5:50 pm

    Feed Rate is how fast the build platform moves when the extruder is ON. Travel Feed Rate is how fast the build platform moves when the extruder is OFF.

    I usually have my travel rate set around 50mm/s, unless I’m printing at a feed rate faster than that, which isn’t very often. Setting the feed rate and travel rate to very different numbers can increase backlash-type noise in your prints and degrade quality.

    People might want to use care if they’re modifying other values in skeinforge – Bre’s info is indeed very good (TY, Bre!) but just plugging that value into the feed rate may only work until you start playing with things like layer heights and thread widths. (that is unless MBI made some serious changes to skeinforge, which they may have)

  • Bobson
    December 22, 2010 at 8:58 pm

    I tried to do this extruder testing thing. But feel kinda inaccurate about my measurement for all things happen in like 3~4 seconds. So, instead of doing it 10 times to get the average, I let the extruding run for 30 seconds. I measured the total length of the extruded string (950mm in my case), and divided it by 30 sec. I think the parameters I’ve got works good on my bot. Always happy to find a rational way tweaking my bot, thanks for the tips!

  • MTO
    December 22, 2010 at 10:48 pm

    @bobson: good idea! That eliminates most of the start/stop error, so you would have a more accurate number…

  • Graphmastur
    December 22, 2010 at 11:20 pm

    The question is, can it do it automatically? 😀

    Good post, bre!

  • Ryanp
    December 22, 2010 at 11:43 pm

    I’ve been wanting to look at this for the last couple of weeks. My approach was going to be to mark a measured length of the filament going in and time how long it takes to go into the top of the extruder. This way you already know the thickness, 3mm and can figure out most of the other skeinforge settings. If the thickness coming out is reliable you can measure that with a micrometer for the same affect.

  • Nick Ames
    December 23, 2010 at 12:26 am

    The thing you are calculating here is the FLOWRATE, not the feedrate. The feedrate is the speed of the platform, and does not need to be measured.

  • Dave Durant
    December 23, 2010 at 1:06 am

    @Nick, I think it just works out right with the profiles they ship with repg. That’s why I was saying earlier that if somebody messes with the other big settings in there, this neat trick probably isn’t going to work any more.

    For the other settings they’ve got on the ToM profile, a feed rate of 25mm/s is correct. They’re just dropping or raising that based on the performance of the DC motor. Sortofa different way to think about things but I don’t doubt it works.

    I wish they’d work on turning up the resolution instead of turning it down, though. 0.40mm layer height? Ick.. Calculate the thread cross section area (layer height * (layer height * wOverT); volume/s = that * feed rate) and work out new layer height and w/t numbers for .30 or .20 instead!! Volume’s volume, once it exits the nozzle.

  • Nicolas Noble
    December 23, 2010 at 2:11 am

    I like the idea of measuring the length of plastic extruded after 30s. This could be done using a piece of gcode calibration script, that would let the plastic go from a fixed height, while moving the platform in a spiral.

  • The Ruttmeister
    December 23, 2010 at 12:56 pm


    Re: higher resolution… does seem a backwards step, but I guess it means faster prints and an easier time getting a good general profile. And they want to be sure people can get good prints from day one.

    But volume isnt always volume… I can’t get below 0.25 layers with PLA, it extrudes so fast that it swells up either side of the nozzel! I need speeds of more than 60mm/s (and thats beyond what my cupcake can manage!

  • Dave Durant
    December 23, 2010 at 1:52 pm

    @Ruttmeister: What’s the volume of, say, a solid 2cm cube printed with a 0.4mm layer height compared to the volume of that same cube printed with a 0.2mm layer height?

    Print time is largely about volume extruded. Where it does change vs resolution is in the travel time – when things are moving and the extruder is off – and the time difference is mostly the difference in the sum of the travel times. For complicated objects with the extruder cycling on & off a lot, this can add up. For more simple objects, it’s not that bad. Going from .4mm to .3mm shouldn’t be bad at all.

    As for PLA, it doesn’t look like repg23 includes any PLA profiles – all I see is ABS. That said, as you drop the layer height, are you just turning up the feed rate or are you also increasing the w/t values? If you start with a profile that works well and work out the thread cross section area, you can use that area to come up with different height & w/t values and not have to mess with the feed & flow rates. There are limits to that, especially with DC extruders, but it works pretty well..

  • Gian Pablo
    January 3, 2011 at 5:06 pm

    This tip alone resolved some niggling print quality issues I’d been seeing (randomly distributed mini blobs).


  • More linkage « Metro State 3D Fab
    January 17, 2011 at 12:20 pm

    […] Skeinforge pt1 Calibrating Skeinforge pt2 Calibrating Skeinforge pt3 Calibrating Skeinforge pt4 Calculating your feedrate PID controller […]

  • Wrapping things up « Metro State 3D Fab
    February 4, 2011 at 1:03 am

    […] we can put all of it together to actually print things. Remember here’s some PID settings and a post to calibrate feed rate. Might be useful. Sorry guys, but no one prints at this point until the software team gives us the […]

  • Printing … with science! - MakerBot Industries
    February 16, 2011 at 12:42 am

    […] Calculating your Feedrate to get awesome prints by Bre Pettis […]

  • fred
    April 4, 2011 at 1:01 pm

    As Nick said, isn’t FEED RATE the speed at which the platform moves, and FLOW RATE the speed at which the extruder pushes filament into the hot end?

    It sounds like you’re trying to set your feed rate to match your measured flow rate, with the idea that pushing 5mm out of the hot end means the platform needs to move 5mm. Do I have that correct?

  • Robert
    February 8, 2014 at 12:48 am

    I would bet my last shirt on the fact that Makerbot uses stepper motors, not DC motors.


Leave your comment


xhtml: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>


What can we help you with today?
I want to chat with Sales.
I have a question about an existing order.
I have a technical question about my device.
Existing Orders
For faster service, enter your order number
(found in your confirmation e-mail)